United States

Big data offers big rewards for manufacturers


The manufacturing industry has always been a leader and innovator in using its data to drive operations.  Over the years, organizations have leveraged their data to enhance quality control, drive efficiencies in the production process, streamline distribution, maximize inventory control and analyze the sales process. Additionally, things such as operational, logistic and executive dashboards have become standard fare. 

As a result, it makes sense that leading manufacturing companies would look to the latest tools and techniques to further leverage their wealth of organizational data. From production data and back-office systems to logistics and sales, organizations are producing an ever-increasing mountain of data. The question is how to make full use of this information. The answer may lie in the concept of big data.

What is big data?

There are a variety of definitions for big data in the industry today. They range from focusing on predictive analytics to fully leveraging unstructured data, such as images, videos and social media, to providing real-time analytics. Big data encompasses the ability to leverage all of an organization's data and can be defined as:

Big data is the set of capabilities enabled by applying advanced statistical and technological techniques and tools to utilize complex, high-volume or disparate data (both structured and unstructured) for decision-making, analysis and reporting.

What drives organizations to venture into the world of big data? It's when they can no longer use traditional methods to fully leverage their data. This usually is due to one or more of the three 'V's

Volume refers to the amount of data produced by an organization. Manufacturing companies produce a wealth of data from all parts of the organization. This includes a significant amount from manufacturing process, distribution, logistics, sales, inventory, finance, etc.

Variety refers to the different types of data an organization produces, including both structured and unstructured data. Manufacturing organizations have a wide range of data at their disposal for use in decision-making, planning and operations, including data from the plant floor, distribution channels, the sales force and inventory systems, to name but a few.

Velocity refers to the speed in which data is produced, as well as the rate at which it is possible to consume it for business purposes. Manufacturing has been using data in a nearly real-time manner for many years in areas such as product quality control and just-in-time inventory management.

It's not necessarily any one of the above (or even the combination of the above) that causes the need for big data capabilities, but rather the complexity that is generated as a result. When data becomes too complex for traditional business intelligence or analytics tools and technologies to process it in a timely manner, big data capabilities are potentially the answer.

Use of big data

Most manufacturing organizations have not come close to leveraging the wealth of data available to them. Big data offers the ability to do so, offering several potential applications, including, but not limited to:

Improved quality control
In most manufacturing companies, the majority of cost is in material and labor. Quality issues in the production process can not only cause lost time and material, but also lost sales, reduced customer confidence and even employee morale issues. Big data can be used to provide real-time feedback and corrective action recommendations related to quality issues. With the volume of data being collected from the production process, more precise quality standards can be defined, tracked and monitored. Big data can even take production quality control to the next level by accurately predicting when a piece of machinery might break or a particular manufacturing step might fail quality control standards. This allows companies to move from reactive to proactive management of quality control.

Operational efficiency
Manufacturing operations can be a complex ecosystem, with many moving parts and many departments performing very different tasks. Understanding the myriad of dependencies and their impacts is often outside of what can be expected from traditional systems, or even human understanding and intuition. Big data is designed to take data from disparate systems and processes and provide insights not normally available through traditional means (potentially in a real-time manner). The applications of these insights into the manufacturing operations are endless. They can be used for:

  • Implementing a higher degree of automation into the manufacturing process
  • Improving scheduling of resources based on anticipated production needs
  • Improving the forecasting of material variety and usage, allowing for improved inventory management
  • Determining areas of potential leakage in the manufacturing process
  • Custom tailoring the manufacturing process to anticipated customer consumption, based on predictive models

Research and development (R&D)
Many manufacturing organizations have internal R&D departments, which are often a prime candidate for the use of big data capabilities. Applications can include:

  • Using social media to drive product improvements
  • Enhancing customer service by using predictive models to determine optimal service and maintenance activities
  • Using customer and operational data to identify potential next-generation products

Supply chain optimization
As the manufacturing process transforms raw materials to products and delivers those products to customers, big data capabilities can be used to streamline work flow. Applications can include:

  • Improving inventory management through better forecasting, procurement management and consumption monitoring
  • Increasing productivity through refined resource management and incorporation of sales forecasts
  • Reducing waste and leakage through improved quality controls and monitoring of the manufacturing process
  • Improving distribution management through predictive models that not only consider internal factors, but also external factors, such as weather forecasts, real-time traffic updates and customer preferences

Sales analysis
Big data can be used to drive improvements in the overall sales process. Applications can include:

  • Using internal and external data (such as social media and customer feedback) to predict customer consumption trends
  • Driving the identification of potential new markets and the probability of success in each
  • Improving segmentation to tailor product offerings to individual customers

There are countless options that big data offers your organization to leverage your data to drive your operations.

Obstacles to the adoption of big data

With their history of being leaders and innovators in using data, why haven't more manufacturing companies ventured into big data?

It's a paradigm shift
Although the concept of big data has been around for several years, most organizations (regardless of industry) have not adopted it as part of the core capabilities. It's a new way of thinking on how to use organizational data, and it usually requires new and cutting-edge technology. It can often result in re-engineering business processes and the way in which people perform their jobs. In a nutshell, it can be a major paradigm shift for an organization. Proper education, planning and training can make the transition into big data easier and less threatening.

Determining the best way to use big data
Being on the cutting edge of new uses of data and the latest application of technology can drive a competitive advantage for organizations, but it can also have the opposite effect. While big data capabilities offer tremendous value, care must be taken to ensure an organization knows why it's using them. Having a strategy, plan and defined success criteria for big data is an important first step and allows for an effective implementation.

Breaking down data silos
Manufacturing organizations have a wealth of data from numerous systems and sources (both internal and external). Unfortunately, this also sometimes comes with the data being managed in silos. Departments can be protective of their data and how it's used. By educating people on what big data is, how the data is used and the resulting value, data silos should start to fade away. Additionally, having an enterprise view of data and how it is managed (i.e. data governance) also facilitates the adoption of big data.

Mistrust of the 'black box'
Because of the way in which processing big data works, there isn't always transparency into how the results were generated. Because of the complexity generated by the volume, variety and velocity of the data processed, there are often different algorithms and statistical techniques applied than in the past. People need to be educated on how the data is being used and processed, as well as shown the validity of the results. Once trust in big data is gained, the use and adoption will increase.

Cost of adoption
Many equate big data with big cost. While full adoption and use of big data capabilities can require a significant investment, it needs to be looked it more holistically within the total cost of ownership equation. Big data can result in more streamlined operations, reduced materials and labor consumption, improved inventory control, more efficient sales and distribution models and an overall improved supply chain. All of these can result in increased revenues and reduced costs.

Having a strategy, plan and understanding of how big data will be used will facilitate implementing it in the most efficient manner. Factor in that big data can provide near real-time insights that traditionally might take weeks or months (or even never) to achieve, and the benefits should more than outweigh the costs.

Additionally, venturing into big data does not necessarily mean a huge upfront cost. Conducting a focused prototype allows you to dip your toe into the big data world to see if it's right for your organization.

In an ever-increasing global competitive landscape, manufacturing companies need to use all the tools at their disposal to create a competitive advantage. One of the most important assets organizations have (and often the most underused) is their data. Most companies are only using a fraction of their data and, even then, do not use it to its full potential. Big data offers the ability to evolve the organization in almost every area.

Developing a strategy around the use of big data is the first step. Understanding how and where it will be used allows for a more efficient and effective implementation. Furthermore, getting started does not need to require a large upfront investment. Conducting a prototype to demonstrate the advantages is a great way to generate support for the use of big data capabilities. Doing nothing puts organizations at risk of continuing with suboptimal operations and even losing market share.